/*
	Estimates the degree of selection bias into the Glassdoor dataset for
	Italian college graduates employed in Italy
	Output:	Figure C1f
			Selection_Italy.csv
*/

local seed "C:\Users\jsock\Dropbox\Research\GD\International"

local dataPath "`seed'/Data"
local inputPath "`seed'/InputData"
local figurePath "`seed'/Replication/Figures"
local tablePath "`seed'/Replication/Tables"
local estimatePath "`seed'/Replication/Estimates"
local tempPath "`seed'/Replication/TempData"

********************************************************
* Save exchange rates as dta file
********************************************************

clear 
set more off
set matsize 10000

insheet using "`inputPath'/Exchange_rates_2022.csv", comma

keep country_glassdoor year ppp_xrat  

save `tempPath'/Exchange_rates.dta, replace

*-------------------------------- 
* Thresholds used for sample selection
*--------------------------------
.
scalar country_premia_thresh = 25

scalar selection_thresh = 25
 
scalar school_thresh = 25

local currentCountry = "Italy" 
local saveCountry = "Italy"

********************************************************
* Read in Australian undergraduate earnings
********************************************************

clear 
set more off
set matsize 10000

insheet using "`inputPath'/`currentCountry'_universities_earnings.csv", comma

keep if year_bach != .

generate country_glassdoor = "`currentCountry'"

merge m:1 country_glassdoor year using `tempPath'/Exchange_rates.dta

keep if _merge == 3
drop _merge

sort school year

rename year_bach undergrad_median_pay
rename year_ms grad_median_pay_1yr
rename v6 grad_median_pay_3yr
rename v7 grad_median_pay_5yr

* Convert to annual gross euros (from monthly euros net taxes)
foreach my_var of varlist *_median_pay*{
	replace `my_var' = `my_var' * 1.25 * 12
}

* Convert to real dollars from nominal euros
foreach my_var of varlist *_median_pay*{
	replace `my_var' = `my_var' * ppp_xrat
}


sort school year

collapse (mean) *_median_pay*, by(school)

save `tempPath'/`currentCountry'_universities_earnings_merge.dta, replace

********************************************************
* Read in country gdp to get universitycountry gdp
********************************************************

clear 
set more off
set matsize 10000
set scheme s1mono

* Set path and load data 
insheet using "`dataPath'/Salaries_international_dataset_main.csv", comma
drop v1

drop if jobtitle == ""

drop metro shortname 
drop city basecurrency country_iso dateval  
drop gender
drop sectorname iscurrentjobflag 

generate national_rank_pct = national_rank / numberuniversities

*--------------------------------
* Exclude users that leave more than 10 reviews
*--------------------------------

sort fk_userid yearofsalary 

by fk_userid : gen userReviews = _N

drop if userReviews > 10

drop userReviews

*--------------------------------
* Convert with ppp exchange rates
*--------------------------------

generate logbase = ln(basesalary * ppp_xrat)

drop if logbase == .

*--------------------------------
* Exclude outliers in base pay
*--------------------------------

scalar scalingThresh = 10

generate realbase = basesalary * ppp_xrat
generate outside_thresh =  (realbase < (1/scalingThresh) * gdppw) | (realbase > scalingThresh * gdppw) 
drop realbase

drop if outside_thresh

*--------------------------------
* Keep only Australia
*--------------------------------

drop if employertypecode == "SELF_EMPLOYED" 

keep if countryname == "`currentCountry'"

keep if universitycountry == "`currentCountry'"

*--------------------------------
* Generate indicators for valid in education analysis
*--------------------------------

generate hasDegree = degree != "UNMATCHED" & degree != "missing" & degree != ""  & degree != "HIGHSCHOOL"

generate uniDegree = degree == "BACHELORS" 

generate hasSchool = school != ""

generate valid_educ = uniDegree & hasDegree & hasSchool & universitycountry != ""

keep if valid_educ 

********************************************************
* Create valid sampels for one and two year comparisons
********************************************************

*--------------------------------
* Years since school ended
*--------------------------------

split endschool, p("-")

destring endschool1, gen(end_year)

generate years_from_graduate = yearofsalary - end_year

*--------------------------------
* Years since school ended
*--------------------------------

bys school : egen avg_grad = median(logbase) if inrange(years_from_graduate,0,1)

bys school : gen n_grad = sum(inrange(years_from_graduate,0,1))

*--------------------------------
* Add comparison earnings
*--------------------------------

generate year = end_year

merge m:1 school using `tempPath'/`currentCountry'_universities_earnings_merge.dta

*--------------------------------
* Comparet External data and GD data for recent graduates
*--------------------------------

preserve

	collapse (max) avg_grad n_grad undergrad_median_pay national_rank_pct, by(school)
	
	generate avg_median_undergrad = ln(undergrad_median_pay)
		
	* Difference between scorecard and gd 
	generate diff_grad = avg_grad - avg_median_undergrad	
	
	egen total_grad = sum(n_grad)
	
	keep if diff_grad != .
	
	* Get averages
	sum diff_grad
	local avg_equal = r(mean)
	scalar est_wtd_equal = r(mean)
	
	sum diff_grad [aw=n_grad]
	local avg_wtd_gd = r(mean)
	scalar est_wtd_gd = r(mean)
	scalar n_wt_wtd_gd = r(N)
	scalar sum_wt_wtd_gd = r(sum_w)

	* Get averages for top 5%
	sum diff_grad if national_rank_pct <= 0.05 [aw=n_grad]
	local avg_wtd_gd_top5 = r(mean)
	scalar est_wtd_gd_top5 = r(mean)
	scalar n_wt_wtd_gd_top5 = r(N)
	scalar sum_wt_wtd_gd_top5 = r(sum_w)
	
	* Get averages for not top 5%
	sum diff_grad if national_rank_pct > 0.05 | national_rank_pct == . [aw=n_grad]
	local avg_wtd_gd_not5 = r(mean)
	scalar est_wtd_gd_not5 = r(mean)
	scalar n_wt_wtd_gd_not5 = r(N)
	scalar sum_wt_wtd_gd_not5 = r(sum_w)

	* Weighted and unweighted kernel density
	twoway kdensity diff_grad [aw=n_grad], lcolor(gs3) lpattern(dash) xscale(r(-1.2(0.4)1.2)) xlabel(-1.2(0.4)1.2)  ///
		xline(0, lcolor(black) lpattern(solid) lwidth(thin)) fcolor(none) ///
		xtitle("log difference within college") ytitle("density") title("") scale(1.2)
	graph export "`figurePath'/Figure_C1f.eps"	, replace

restore

*--------------------------------
* Export summary of selection estimates
*--------------------------------

preserve

	generate selection_est_equal = est_wtd_equal
	generate selection_est_wtd = est_wtd_gd
	generate selection_est_wtd_top5 = est_wtd_gd_top5
	generate selection_est_wtd_not5 = est_wtd_gd_not5

	generate selection_wtd_N = n_wt_wtd_gd
	generate selection_wtd_N_top5 = n_wt_wtd_gd_top5
	generate selection_wtd_N_not5 = n_wt_wtd_gd_not5

	generate selection_wtd_sum = sum_wt_wtd_gd
	generate selection_wtd_sum_top5 = sum_wt_wtd_gd_top5
	generate selection_wtd_sum_not5 = sum_wt_wtd_gd_not5

	keep universitycountry selection_*
	
	keep if _n == 1
	
	order universitycountry selection_wtd_N* selection_wtd_sum* selection_est_* 
	
	outsheet using "`estimatePath'\Selection_`saveCountry'.csv" , comma replace

restore
